Transport in isolated rat hepatocytes of the phospho - Europe PMC

Biochem. J. (1991) 274, 369-374 (Printed in Great Britain)

369

Transport in isolated rat hepatocytes of the phosphooligosaccharide that mimics insulin action Effects of adrenalectomy and glucocorticoid treatment Jose F. ALVAREZ,* Juan Antonio SANCHEZ-ARIAS,t Ana GUADANO,* Francisco ESTEVEZ,* Isabel VARELA,* Juan E. FELIUt and Jose M. MATO*t *Instituto de Investigaciones Biomedicas, C.S.I.C., Arturo Duperier 4, 28029-Madrid, and

tServicio de Endocrinologia Experimental, Hospital Puerta de Hierro, Universidad Aut6noma de Madrid,

San Martin de Porres 4, 28035-Madrid, Spain

The addition to intact cells of an inositol phospho-oligosaccharide (POS), which is the polar head-group of an insulinsensitive glycosylphosphatidylinositol, mimics and may mediate some of the biological effects of this hormone. Here we report the existence of a POS transport system in hepatocytes. This POS transport system is specific and time- and dosedependent. Insulin-resistance caused by dexamethasone administration to rats was accompanied by a decrease in the hepatocyte POS transport system. In contrast, bilateral adrenalectomy provoked a significant increase in the transport of POS. Both the temporal uptake of POS and the regulation of this process by conditions known to modify the sensitivity to insulin suggest that this novel transport system might be involved in the insulin signalling mechanism.

INTRODUCTION

extracellular POS can be transported into the cells. In this paper give evidence for the existence of a POS transport system in isolated rat hepatocytes.

we

In the last few years, a glycosylphosphatidylinositol (glycosylPI) has been implicated in insulin action [1-4]. Insulin-dependent glycosyl-PI hydrolysis has been reported in BC3H1 myocytes [2], H35 hepatoma cells [3], rat hepatocytes [5] and T-lymphocytes [6]. The polar head-group of this glycosyl-PI, a phosphooligosaccharide (POS) containing a glucosamine-inositol-phosphate moiety, galactose and several additional phosphates [7], has been reported to mimic some of the biological actions of insulin when added to intact adipocytes or hepatocytes. POS inhibits isoprenaline-stimulated phospholipid methyltransferase [8] and lipolysis [9], as well as stimulating lipogenesis [10] in adipocytes; furthermore, it antagonizes the effects of glucagon on glycogen phosphorylase a and pyruvate kinase activities and on cyclic AMP levels in isolated rat hepatocytes [11]. Moreover, POS has been found to evoke the phosphorylation and dephosphorylation of some of the same cellular phosphoproteins as does insulin [4, 12], and to stimulate amino acid uptake [13]. In addition, this molecule has been reported to modulate the activity of certain enzymes when added to cell extracts or to the purified enzymes. POS stimulates cyclic AMP phosphodiesterase [4] and pyruvate dehydrogenase [14-16], and inhibits adipocyte adenylate cyclase [14] and bovine heart cyclic AMP-dependent protein kinase [16,17]. A variety of experiments indicate that POS is generated extracellularly. The majority (about 80 %) of the insulin-sensitive glycosyl-PI in rat hepatocytes is localized at the outer cell surface [5,18], the extracellular levels of POS have been shown to increase in BC3HI myocytes after the addition of insulin [19], and Reuber hepatoma cells release constitutively POS-like substances into the culture media [20]. Moreover, the addition to intact BC3H1 myocytes of an antibody raised against the oligosaccharide moiety of glycosyl-PI-anchored membrane proteins blocks the insulin-induced stimulation of pyruvate dehydrogenase and the effects of POS in vitro [21]. However, it is not known whether

Abbreviations used: PI, phosphatidylinositol; POS, t To whom correspondence should be addressed.

Vol. 274

EXPERIMENTAL Materials NaB3H4 (15 Ci/mmol) was from New England Nuclear, and [U-14C]glucose (280 mCi/mmol) was purchased from Amersham. PI-specific phospholipase C purified from Bacillus thuringiensis [22] was generously supplied by Dr. S. Udenfriend, Roche Institute of Molecular Biology, Nutley, NJ, U.S.A. Galactose oxidase from Doctylium dendroides, ,-galactosidase from Escherichia coli and lipid standards were purchased from Sigma. Silica gel G plates for t.l.c. were from Scharlau. Collagenase was from Boehringer Mannheim and human insulin (Actrapid HM) was obtained from Novo Industri A/S. Dexamethasone phosphate (Fortecortin) was obtained from Merck. All the remaining reagents were of analytical grade. Labelling of purified glycosyl-PI with galactose oxidase/NaB3H4 and generation of 13Hlgalactose-labelled POS Glycosyl-PI was purified from rat liver membranes by sequential t.l.c. as previously described [3,7]. Purified glycosyl-PI from 20 rat livers was resuspended, using sonication, into 0.3 ml of 50 mM-phosphate buffer, pH 8.0, and treated with galactose oxidase (5 units) and 2 mCi of NaB3H4 for 1 h at 37 °C [23]. The mixture was then dried under vacuum, dissolved in chloroform/methanol (2: 1, v/v) and purified by t.l.c. on silica gel G [3]. The plate was developed twice in chloroform/acetone/methanol/ glacial acetic acid/water (10:4:2:2:1, by vol.). A 1 cm region around the origin was eluted with 2 x 2 ml of methanol at 37 °C, and this fraction was re-chromatographed in chloroform/ methanol/NH4OH/water (45:45:4: 10, by vol.). Regions of I cm were then scraped off, and lipids were eluted with 2 x 2 ml of methanol at 37 °C and the radioactivity associated with each

phospho-oligosaccharide.

J. F. Alvarez and others

370 fraction was determined by counting a sample for radioactivity. Labelled glycosyl-PI was then dissolved into 0.75 ml of chloroform/methanol (2: 1, v/v) containing 0.03 M-HCI, and 0.2 ml of water was added to form two phases. After vigorous shaking the organic phase was removed and retained; the water phase was washed once with 0.3 ml of chloroform/methanol (2: 1, v/v) and the organic phases were combined. The efficacy ofthe purification procedure was assessed by h.p.l.c. of the glycosyl-PI labelled with NaB3H4 using a silica column (Ultrasil-Si, 10 ,tm pore size, 4.6 mm x 250 mm; Beckman). A 20 min linear gradient from chloroform/methanol/glacial acetic acid (14:2:1, by vol.) to chloroform/methanol/glacial acetic acid/water (40:45:10:2, by vol.) was used. The flow rate was 1 ml/min and the elution of the labelled glycosyl-PI was detected by measuring the amount of radioactivity in each fraction. Purified labelled glycosyl-PI was evaporated to dryness under N2, resuspended by sonication into 0.2 ml of 20 mM-sodium borate, pH 7.4, and treated with 1 unit of PI-specific phospholipase C (1 unit cleaves 0.8 nmol of PI/min at 37 °C [22]) for 12 h at 37 °C to generate labelled POS. The reaction was stopped by the addition of 0.75 ml of chloroform/ methanol (2: 1, v/v) containing 0.03 M-HCl to form two phases. After vigorous shaking, the water phase was removed and retained; the organic phase was washed once with 0.5 ml of 5 mm-NaCl in 50 % methanol and the water phases were combined and lyophilized. After lyophilization, the sample was dissolved in water and the specific radioactivity of the labelled POS was adjusted to approx. 25000 d.p.m./nmol by the addition of non-labelled POS. The amount of POS was determined by measuring the concentration of free amino groups with fluorescamine [24]. Non-labelled POS was obtained by reaction of purified non-labelled glycosyl-PI with PI-specific phospholipase C as described above. The biological activity of POS thus generated was assessed by determining its ability to inhibit cyclic AMP-dependent protein kinase [17] and to stimulate amino acid transport in isolated rat hepatocytes [18]. This assay with intact cells was preferred to previous ones [9,111, since it measures a stimulatory rather than an inhibitory effect of POS. Hepatocyte isolation and measurement of POS uptake Fed male Wistar rats (180-200 g) from our inbred colony were used. Some of the animals were subjected to bilateral adrenalectomy, or to simulated operation, under Pentothal anaesthesia 5 days before they were used; adrenalectomized rats received 0.45 % (w/v) NaCl solution as drinking water. Treatment of rats with glucocorticoids was carried out by administration of three doses of dexamethasone phosphate (0.2 mg/l00 g body wt., subcutaneously) at 50, 26 and 2 h before the isolation of the hepatocytes [25]. Hepatocytes were isolated by perfusion of the liver with collagenase [26]. Cells were suspended in KrebsHenseleit medium in the presence of 10 mM-glucose and incubated as reported elsewhere [11]. The viability of the isolated hepatocytes was evaluated by the Trypan Blue test; usually 90-95 % of the cells excluded the stain. Unless otherwise stated, after 30 min preincubation [3H]POS (10 /LM final concentration, about 50000 d.p.m.) was added to cell suspensions (200 #1 final volume) and incubated at 37 'C. At various times, reactions were stopped by the addition of I ml of ice-cold Krebs-Henseleit medium and without delay centrifuged for 10 s at 2000 rev./min in a Microfuge. The cell pellet was immediately resuspended and washed with 2 x 1 ml of ice-cold Krebs-Henseleit buffer. After the last centrifugation, the pellet was resuspended into 250 ,ul of water, and 250 ,ul of 20 % (w/v) trichloroacetic acid was added. After standing for 20 min at 4 'C, samples were centrifuged and the radioactivity present in the supernatant was determined. Results are expressed as pmol of POS associated with the cell pellet/mg of protein. To determine glycogen synthesis, [U-

14C]glucose was added to the hepatocyte suspension to a final concentration of 20 mm. After 30 min, samples were taken for the measurement of [14C]glucose incorporation into glycogen [25]. Protein was determined by the method of Bradford [27]. All assays were carried out in triplicate. The statistical significance of differences between values was calculated by Student's t test. The differences were considered to be statistically significant when the P value was less than 0.05. RESULTS Radiolabelling of glycosyl-PI with galactose oxidase and

NaB3H4 Glycosyl-PI was purified from rat liver microsomes and labelled by treatment with galactose oxidase and NaB3H4 as described in the Experimental section. After labelling, radioactive glycosyl-PI was purified by sequential t.l.c. As shown in Fig. 1, one major radioactive peak, which comigrated with authentic 60oI




CD c O

' 0

Et o0

2

I

4-

-10a0-9 I

-8

log{[Insulin] (M)} Fig. 7. Effect of insulin on the rate of glycogen synthesis Glycogen synthesis was measured in hepatocytes isolated from sham-operated (E, *) and adrenalectomized (0, 0) rats, treated (O, 0) or not (El, 0) with dexamethasone. Values are the means+S.E.M. of six to nine experiments.

Table 3. Effects of dexamethasone treatment on insulin- and POSstimulated glycogen synthesis in isolated rat hepatocytes

Glycogen synthesis was determined as described in the Experimental section. Values are the means +S.E.M. of three experiments. The effect of dexamethasone treatment on insulin or POS-stimulated glycogen synthesis was statistically significant compared with that in sham-operated animals (P < 0.05). Glycogen synthesis (%)

Control rats Sham-operated rats Sham-operated rats+ dexamethasone

Basal

POS (10 ,UM)

Insulin (10 nM)

100 100 100

153 +8 168+4 86+ 5

187 +9 172+20 105+9

Table 4. Specificity of the uptake of POS by hepatocytes

Hepatocytes were resuspended in Krebs-Henseleit medium in the presence of 10 /SM-[galactose-3H]POS and the various carbohydrates mentioned in the Table (100 /ZM final concentration). After a 5 min incubation, the amount of radioactivity associated with the cells was measured as described in the Experimental section. The 100 % value was 2723 + 298 d.p.m. of [3H]POS/mg of protein. The means + S.E.M. of three to four experiments are presented. Additions (100 /LM) None

myo-Inositol

Inositol 1-phosphate Inositol 2-phosphate Inositol hexakisphosphate Mannose Galactose Glucosamine

Vol. 274

POS uptake (%) 100 90+ 8 81+6 82+12 57+ 13 67 + 11 63 + 5 45 +6

DISCUSSION Insulin has been shown to promote the hydrolysis of a novel glycosyl-PI (reviewed in [28,29]). The polar head-group of this glycosyl-PI is a POS that contains a glucosamine-inositolphosphate moiety, about four galactosyl residues and several additional phosphate groups which are probably associated with a galactosyl residue [7,24]. The presence of a cellular uptake process for POS has been suggested by the observation that the addition of this molecule to intact adipocytes or hepatocytes mimics certain biological effects of insulin [4,8-12], by the finding that POS is released into the incubation media of BC3H1 [19] and Reuber hepatoma cells [20], by the localization of the majority of the glycosyl-PI precursor at the outer surface of hepatocytes [5,18], and by the use of antibodies raised against the oligosaccharide moiety of glycosyl-PI-anchored proteins to block insulin action [21]. To investigate the mechanism by which POS acts in intact cells, we have developed a procedure to generate [galactose3H]POS by radiolabelling the glycosyl-PI precursor with NaB3H4 and galactose oxidase, followed by hydrolysis with a PI-specific phospholipase C. Using [galactose-3H]POS thus prepared, we observed an efficient transport of POS by isolated rat hepatocytes. Evidence for this conclusion comes from the finding that POS uptake was detected at 37 'C but not at 4 'C, and from the observation that after being taken up by the cells radioactive POS was not released into the incubation medium either by incubation in the absence of POS or by the addition of non-labelled POS. Cellular accumulation of POS was abolished by metabolic poisoning with KCN. These results indicate that the uptake of POS is probably an energy-dependent process and cannot be explained by diffusion ofthe molecule through the cell membrane. The uptake of POS was time- and dose-dependent. Uptake was observed within 5 min of incubation, which conforms with the biological effects of POS in intact adipocytes or hepatocytes; these are observed within 5-10 min of its addition [4,11], and with the insulin-dependent hydrolysis of glycosyl-PI in myocytes [2], H35 hepatoma cells [3] and hepatocytes [5], which is maximal within 2-5 min of the addition of the hormone. The data of Fig. 5 do not permit us to estimate the affinity of the transport system. Since the uptake system is not yet saturated at 1 mM-POS, the affinity seems to be low relative to the likely physiological concentrations of POS. The concentration of POS generated in response to insulin is not known. Studies thus far suggest that the majority of the glycosyl-PI precursor of POS is present at the outer cell surface [4,18]. The organization and dynamic state of glycosyl-PI molecules is not known, although aggregation of certain glycolipids in lipid bilayers has been reported [30]. If glycosyl-PI molecules are concentrated in certain regions of the biomembrane and transport of POS occurs close to its site of production, the concentration of POS at this site might be higher than predicted for a molecule whose lipid precursor accounts for about 0.5 % of the total plasma membrane phospholipids [5]. The data presented here indicate that inositol hexakisphosphate, mannose, galactose and glucosamine all inhibited

374 POS uptake to a similar extent (about 40-50 %). It has been previously reported that, in adipocytes, inositol monophosphate, mannose and glucosamine do not alter the effects of insulin or POS on isoprenaline-stimulated cyclic AMP levels or lipolysis [31]. Similarly, in hepatocytes, mannose, glucosamine, galactose, myo-inositol and inositol hexakisphosphate have no effect on POS- or insulin-stimulated amino acid uptake [13]. In contrast, the stimulatory effect of POS on lipogenesis is blocked by mannose, glucosamine and inositol monophosphate [31]. Similarly, the effects of POS and insulin on protein phosphorylation in adipocytes are blocked by inositol phosphates and glucosamine [32]. These data have been interpreted as being suggestive of the existence of an uptake system for POS [31,32]. The mechanisms by which these sugars specifically affect only some of the biological effects of insulin or POS cannot yet be explained. In man and animals, glucocorticoid excess is associated with insulin-resistance [33,34]. Conversely, adrenalectomized animals are abnormally sensitive to this hormone [35]. The mechanism by which glucocorticoids modulate insulin action is not well understood. There is experimental evidence suggesting that these steroids may influence insulin effects by acting at the insulin receptor level and at a post-receptor step [34]. Previously we have shown that insulin-resistance caused by dexamethasone administration is accompanied by a marked decrease in both the hepatic levels of the insulin-sensitive glycosyl-PI and the number of insulin receptors [25]. In contrast, bilateral adrenalectomy raised the cellular content of glycosyl-PI and the number of insulin receptors [25]. The present results indicate that dexamethasone administration is accompanied by a decrease in the hepatic uptake of POS, and that bilateral adrenalectomy enhances this process. Although the effect of dexamethasone on POS uptake was relatively small (about 30 %), the inhibition of insulin- or POS-stimulated glycogen synthesis was almost complete. These results indicate that these steroids may influence insulin action by acting at more than one level and that the inhibition by dexamethasone of insulin- or POS-stimulated glycogen synthesis cannot be explained only by a decrease in the uptake of POS. In conclusion, although there is no direct evidence for a role of this uptake system for POS in insulin action, the data described here on the temporal uptake of POS, and on the regulation of this process by conditions known to modify sensitivity to insulin, support the idea of such a transport system being functionally important in insulin signalling. J.F.A., J.A.S.-A., F.E., A.G. and I.V. are fellows of the Caja de Ahorros de Madrid, Fondo de Investigaciones Sanitarias, Consejeria de Educaci6n, Cultura y Deportes del Gobierno de Canarias, Europharma and Ministry of Education respectively. This investigation was supported by grants from CICYT, FIS, Europharma and Glaxo.

REFERENCES 1. Saltiel, A. R. & Cuatrecasas, P. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 5793-5797

J. F. Alvarez and others 2. Saltiel, A. R., Fox, J. A., Sherline, P. & Cuatrecasas, P. (1986) Science 233, 967-972 3. Mato, J. M., Kelly, K. L., Abler, A. & Jarett, L. (1987) J. Biol. Chem. 262, 2131-2137 4. Alemany, S., Mato, J. M. & Stralfors, P. (1987) Nature (London) 330, 77-79 5. Alvarez, J. F., Varela, I., Ruiz-Albusac, J. M. & Mato, J. M. (1988) Biochem. Biophys. Res. Commun. 152, 1455-1462 6. Gaulton, G., Kelly, K. L., Mato, J. M. & Jarett, L. (1988) Cell 53, 963-970 7. Mato, J. M., Kelly, K. L., Abler, A., Jarett, L., Corkey, B. E., Cashel, J. A. & Zopf, D. (1987) Biochem. Biophys. Res. Commun. 146, 764-770 8. Kelly, K. L., Mato, J. M. & Jarett, L. (1986) FEBS Lett. 209, 238-242 9. Kelly, K. L., Mato, J. M., Merida, I. & Jarett, L. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 6404-6407 10. Saltiel, A. R. & Sorbara-Cazan, L. R. (1987) Biochem. Biophys. Res. Commun. 149, 1084-1092 11. Alvarez, J. F., Cabello, M. A., Feliu, J. E. & Mato, J. M. (1987) Biochem. Biophys. Res. Commu. 147, 765-771 12. Kelly, K. L., Merida, I., Wong, E. H. A., DiCenzo, D. & Mato, J. M. (1987) J. Biol. Chem. 262, 15282-15290 13. Varela, I., Avila, M., Mato, J. M. & Hue, L. (1990) Biochem. J. 267, 541-544 14. Saltiel, A. R. (1987) Endocrinology (Baltimore) 120, 967-972 15. Gottschalk, K. W. & Jarett, L. (1988) Arch. Biochem. Biophys. 261, 175-185 16. Larner, J., Huang, L. C., Schwartz, C. F. W., Oswald, A. S., Shen, T. Y., Kinter, M., Tang, G. & Zeller, K. (1988) Biochem. Biophys. Res. Commun. 151, 1416-1426 17. Villalba, M., Kelly, K. L. & Mato, J. M. (1988) Biochim. Biophys. Acta 968, 69-76 18. Varela, I., Alvarez, J. F., Clemente, R., Ruiz-Albusac, J. M. & Mato, J. M. (1990) Eur. J. Biochem. 188, 213-218 19. Romero, G., Lutrell, L., Rogol, A., Zeller, K., Hewlett, E. & Larner, J. (1988) Science 240, 509-511 20. Witters, L. A. & Watts, T. D. (1988) J. Biol. Chem. 263, 8027-8032 21. Romero, G., Gimez, G., Huang, L. C., Lilley, K. & Lutrell, L. (1990) Proc. Natl. Acad. Sci. U.S.A. 87, 1476-1480 22. Howard, A. D., Berger, J., Gerber, L., Familletti, P. & Udenfriend, S. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 6055-6059 23. Steck, T. L. & Dawson, G. (1974). J. Biol. Chem. 249, 2135-2142 24. Merida, I., Corrales, F. J., Clemente, R., Ruiz-Albusac, J. M., Villalba, M. & Mato, J. M. (1988) FEBS Lett. 236, 251-255 25. Cabello, M. A., Sanchez-Arias, J. A., Liras, A., Mato, J. M. & Feliu, J. E. (1990) Mol. Cell. Endocrinol. 68, RI-R5 26. Feliu, J. E., Mojena, M., Silvestre, R. A., Monge, L. & Marco, R. (1983) Endocrinology (Baltimore) 112, 2120-2126 27. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 28. Low, M. G. & Saltiel, A. R. (1987) Science 239, 268-275 29. Mato, J. M. (1989) Cell. Signalling 1, 143-146 30. Hakomori, S. I. (1981) Annu. Rev. Biochem. 50, 733-764 31. Machicao, F., Mushack, J., Seffer, E., Ermel, B. & Haring, H. U. (1990) Biochem. J. 266, 909-916 32. Stralfors, P. & Alemany, S. (1990) FEBS Lett. 268, 169-172 33. Rizza, R. A., Mandarino, L. J. & Gerich, J. E. (1982) J. Clin. Endocrinol. Metab. 54, 131-138 34. McMahon, M., Gerich, J. & Rizza, R. A. (1988) Diabetes Metab. Rev. 4, 18-30 35. Wajchenberg, B. L., Pereira, V. G., Pupo, A. A., Shanaider, J., Ulhoa Cintra, A. B. & Mattar, E. (1964) Diabetes 13, 170-175 36. Ferguson, M. A. J., Haldar, K. & Cross, G. A. M. (1985) J. Biol. Chem. 260, 4963-4968

Received 11 April 1990/26 September 1990; accepted 23 October 1990

1991